Beyond Language Models: Byte Models are Digital World Simulators

TL;DR

This paper introduces bGPT, a byte-level model that predicts bytes to simulate the digital world, demonstrating high accuracy in tasks like music conversion and CPU behavior simulation, thus extending deep learning beyond language models.

Contribution

The paper presents bGPT, a novel byte prediction model that effectively simulates digital data and hardware processes, outperforming prior models in multi-modal digital world tasks.

Findings

Achieves 0.0011 bits per byte error in music data conversion

Exceeds 99.99% accuracy in CPU behavior simulation

Performs comparably to specialized models across modalities

Abstract

Traditional deep learning often overlooks bytes, the basic units of the digital world, where all forms of information and operations are encoded and manipulated in binary format. Inspired by the success of next token prediction in natural language processing, we introduce bGPT, a model with next byte prediction to simulate the digital world. bGPT matches specialized models in performance across various modalities, including text, audio, and images, and offers new possibilities for predicting, simulating, and diagnosing algorithm or hardware behaviour. It has almost flawlessly replicated the process of converting symbolic music data, achieving a low error rate of 0.0011 bits per byte in converting ABC notation to MIDI format. In addition, bGPT demonstrates exceptional capabilities in simulating CPU behaviour, with an accuracy exceeding 99.99% in executing various operations. Leveraging next byte prediction, models like bGPT can directly learn from vast binary data, effectively simulating the intricate patterns of the digital world.